Studies in Polish Linguistics vol. 11,12, 2016,2017, issue 3,2, pp. 75–7575–95 10.4467/23005920SPL.17.003.7021doi: 10.4467/23005920SPL.17.004.7022 www.ejournals.eu/SPL Björn Wiemer Johannes Gutenberg University Mainz Anna Socka University of Gdańsk How Much Does Pragmatics Help to Contrast the Meaning of Hearsay Adverbs? (Part 2) Abstract The present study aims at differentiating between semantically-coded and pragmatically- conditioned meaning components of Polish and German sentence adverbs whose meaning is conventionally associated with hearsay (≈ Eng. allegedly, reportedly, supposedly). In the current part of the study, we argue why our objective should be reached on the basis of Generalized Conversational Implicatures (GCIs), and we show which particular commu- nicative principles distinguished in Neo-Gricean frameworks can sensibly be considered as triggers of GCIs that evoke ‘epistemic overtones’ in the use of hearsay adverbs. We dif- ferentiate between GCIs which work for all relevant adverbs and implicatures which only apply to more individual properties of hearsay adverbs on more specific, “deeper” levels of their meaning structure. In accordance with this more descriptive task, we discuss general issues concerning presumable hierarchies of factors that influence (trigger or cancel) epis- temic implicatures in the usage of lexical markers of information source. We argue that many discourse properties on the semantics-pragmatics interface which are characteristic of grammatical evidentials also hold true for lexical markers of information source. Keywords Polish, German, reportive evidentiality, sentence adverbs, Generalized Conversational Im- plicatures, coded vs. inferred meaning Streszczenie Artykuł stanowi próbę rozróżnienia zakodowanych semantycznie oraz uwarunkow- anych pragmatycznie komponentów znaczenia polskich i niemieckich reportatywnych przysłówków zdaniowych (ang. allegedly, reportedly, supposedly). W niniejszej, drugiej części artykułu na podstawie teorii Uogólnionych Implikatur Konwersacyjnych (General- ized Conversational Implicatures, GCI) pokazujemy, w jaki sposób mechanizmy komuni- kacyjne przyjęte w ujęciach neo-Grice’owskich prowadzą do GCI nadających przysłówkom reportatywnym zabarwienie epistemiczne. Rozróżniamy przy tym GCI towarzyszące użyciu wszystkich przysłówków reportatywnych oraz te implikatury, które wiążą się z ich 76 Björn Wiemer, Anna Socka indywidualnymi cechami na głębszym poziomie struktury znaczeniowej. Następnie po- ruszamy problem ogólniejszy, dotyczący przypuszczalnych hierarchii czynników, które wywołują (lub znoszą) implikatury epistemiczne u jednostek leksykalnych wyrażających źródło informacji. Uważamy, że jednostki te wykazują na poziomie dyskursu wiele właściwości dotyczących styku semantyki i pragmatyki, które dotychczas przypisywano tylko gramatycznym eksponentom ewidencjalności. Słowa kluczowe język polski, język niemiecki, ewidencjalność reportatywna, przysłówki sentencjalne, uogól- nione implikatury konwersacyjne, znaczenie zakodowane vs. znaczenie wywnioskowane In the first part of this paper published in the preceding issue of this journal (Wiemer and Socka 2017), we presented a systematic corpus study of hearsay adverbs in Polish and German providing the empirical basis for the conclu- sions that we want to draw in this part. 3. A proposal of how to explain the facts How can we make sense of the facts assembled in Wiemer and Socka (2017: Sec- tion 2.2)? We are first going to justify why we think that the relation between re- portive justification and epistemic judgment can best be characterized as result- ing from Generalized Conversational Implicature. Probably, another name for this property is ‘default’, in the same way as, in Wiemer and Socka (2017: Section 2.2.2), we pointed out origo-exclusiveness as a feature of rzekomo (in distinction to other Polish allegedly-units) that, on the one hand, arises independently of the spe- cific context and, on the other hand, can be canceled by other factors. Apart from this individual feature of rzekomo, we are making the more general claim that only the evidential (reportive) component is inherent to all allegedly-units in Polish and German, while epistemic overtones result from an interaction of this stable (i.e. coded) meaning with some general assumptions about the way communica- tion works (3.1). Then we try to establish, on Neo-Gricean grounds, which spe- cific kind of implicature is responsible for the assumed mechanism (3.2), and we discuss some more far-reaching consequences of this analysis of different kinds of implicatures for the semantic-pragmatics interface of evidential markers (3.3). 3.1. Epistemic overtones as the result of Generalized Conver- sational Implicatures Our own proposal amounts to ascribing the epistemic overtones in the exam- ined units the status of Generalized Conversational Implicatures (henceforth GCI). According to Levinson (2000: 16), GCIs differ from Particularized Con- versational Implicatures in the following way: How much does pragmatics help to contrast the meaning of hearsay adverbs? (Part 2) 77 [1] a) An implicature I from utterance U is particularized iff U implicates I only by virtue of specific contextual assumptions that would not invariably or even nor- mally obtain. b) An implicature I is generalized iff U implicates I unless there are unusual specific contextual assumptions that defeat it. GCIs take place as preferred, or default, interpretations that are canceled (or blocked) only under some specific conditions. GCIs are associated with ut- terance-type meanings, which Levinson postulates to be a third kind of mean- ing because they cannot be reduced to either sentence-type-meaning or utter- ance-token-meaning, but rather belong to “a level of systematic pragmatic inference not based on direct computations about speaker-intentions but on general expectations about how language is normally used” (Levinson 2000: 20; emphasis added, BW/AS). Among other things, the observations made in Wiemer and Socka (2017: Section 2.2) show that, on the one hand, any of the reportive adverbs (a) eas- ily triggers (or: is compatible with) an implicature of negative epistemic assess- ment, but, on the other hand, (b) any of these units allows this implicature to be canceled. Admittedly, the contextual conditions on which this occurs differ. They are obviously most specific for Pol. rzekomo and, probably, Germ. ange- blich. But this difference can be captured by introducing defaults (or, converse- ly, defeatability) of different strength. We could claim that for Pol. rzekomo the default of implicating negative epistemic judgment is stronger than for jakoby and podobno. How ‘epistemic strength’ might be reliably assigned and used as a falsifiable comparative concept on any sort of scale is an issue that, at pre- sent, we do not want, and do not need, to decide upon. What is important here is that news reports about purported violations of laws (and similar kinds of discourse) readily supply contexts that defeat this implicature (= default as- sumption). For the same reason, we can justify why the component of epistemic judg- ment for, e.g., rzekomo cannot be characterized as Conventional Implicature. If it could, we would expect this component to be neither calculable, nor cancela- ble (so that it would have to be stipulated in the adverb’s lexical entry as its in- dividual feature). But as we saw, it can be canceled even in rzekomo, so that this unit basically behaves like the other reportive adverbs, irrespective of whatever strength of default we might wish to assign to it. Apart from that, it has been claimed that conversational implicatures tend to be universal, whereas conventional ones tend not to be universal. This leaves us with the question of how universal (i.e. culture-independent) conversa- tional implicatures characteristic of reportive adverbs (or other propositional modifiers) are, regardless of which kind of principle is responsible for the GCI (see 3.2). We will return to this issue in section 4. 78 Björn Wiemer, Anna Socka 3.2.Now Whichwe are going specific to discuss kind which of GCI mechanism is at work? causes the supposed machin- ery on the lexicon-pragmatics interface to work. For this purpose, we rely on a Neo-Gricean approach defended and summarized in Huang (2007). 3.2.1. Principles based on the Quantity maxim:1 These principles are usually explicated in the following schematic way. Note that they (as well as other principles below) are divided up between speaker and addressee: [2] Speaker: Do not say less than is required (bearing the I-principle in mind). Addressee: What is not said is not the case. (Huang 2007: 41) The Quantity maxim is subdivided into those based on scalar implicatures and on clausal implicatures. We turn to each of these now. 3.2.1.1. Based on Horn-scale (Horn’s Q-principle), or Q-scalar implicatures? Scalar implicatures take the general form [3] Q-scalar : <x, y > 2 y + > Q-scalar ~x (Huang 2007: 42) We see quite quickly that this kind of implicature cannot be responsible for the epistemic overtones of hearsay adverbs. If – on the basis of scalar assump- tions mentioned after Ramat and Ricca (1998) in Wiemer and Socka (2017: Section 2.2) – we assume that the epistemic component (negative stance) con- stitutes a scale and x = allegedly (rzekomo), y = reportedly (podobno), then by using reportedly, the speaker implicates ~allegedly. Or, the other way round, x should entail y. With units of standard Horn-scales this allows one to say,